Assessment of variables associated with prolonged admission duration in children with Mycoplasma pneumoniae pneumonia.

INTRODUCTION: Macrolide-resistant Mycoplasma pneumoniae (MRMP) has become prevalent in children. This study investigated the clinical and laboratory variables of MRMP and macrolide-sensitive M. pneumoniae (MSMP) and identified factors associated with prolonged hospital admission in children. METHODS: A prospective multicenter study was conducted in 1063 children <18 years old in July 2018-June 2020. The 454 had a positive M. pneumoniae polymerase chain reaction assay. RESULTS: Most subjects had MRMP (78.4%), and all mutated strains had the A2063G transition. We defined MRMP* (n = 285) as MRMP pneumonia requiring admission and MSMP* (n = 72) as MSMP pneumonia requiring admission. Patients with MRMP pneumonia were older, more likely to have segmental/lobar pneumonia, and had more febrile days than those with MSMP pneumonia. C-reactive protein (CRP), lactate dehydrogenase (LDH), and percentage neutrophils were more strongly associated with MRMP* than MSMP* groups. Percentage neutrophils, CRP, and alanine aminotransferase significantly changed between admission and follow-up measurements in patients with MRMP* (P < 0.05). The duration of admission positively correlated with the number of febrile days after initiation of antibiotic medication and laboratory variables (white blood cell count, CRP, and aspartate aminotransferase [AST]) (P < 0.05). Random forest analysis indicated that the number of febrile days after initiation of antibiotic medication, AST, and percentage neutrophils at admission was over five. CONCLUSIONS: This study indicated that children with M. pneumoniae pneumonia with a higher number of febrile days after initiation of antibiotic medication, AST, and percentage neutrophils at admission were more likely to have prolonged admission duration.

Optimizing High-Performance Computing Systems for Biomedical Workloads.

The productivity of computational biologists is limited by the speed of their workflows and subsequent overall job throughput. Because most biomedical researchers are focused on better understanding scientific phenomena rather than developing and optimizing code, a computing and data system implemented in an adventitious and/or non-optimized manner can impede the progress of scientific discovery. In our experience, most computational, life-science applications do not generally leverage the full capabilities of high-performance computing, so tuning a system for these applications is especially critical. To optimize a system effectively, systems staff must understand the effects of the applications on the system. Effective stewardship of the system includes an analysis of the impact of the applications on the compute cores, file system, resource manager and queuing policies. The resulting improved system design, and enactment of a sustainability plan, help to enable a long-term resource for productive computational and data science. We present a case study of a typical biomedical computational workload at a leading academic medical center supporting over $100 million per year in computational biology research. Over the past eight years, our high-performance computing system has enabled over 900 biomedical publications in four major areas: genetics and population analysis, gene expression, machine learning, and structural and chemical biology. We have upgraded the system several times in response to trends, actual usage, and user feedback. Major components crucial to this evolution include scheduling structure and policies, memory size, compute type and speed, parallel file system capabilities, and deployment of cloud technologies. We evolved a 70 teraflop machine to a 1.4 petaflop machine in seven years and grew our user base nearly 10-fold. For long-term stability and sustainability, we established a chargeback fee structure. Our overarching guiding principle for each progression has been to increase scientific throughput and enable enhanced scientific fidelity with minimal impact to existing user workflows or code. This highly-constrained system optimization has presented unique challenges, leading us to adopt new approaches to provide constructive pathways forward. We share our practical strategies resulting from our ongoing growth and assessments.

Big Omics Data Experience.

As personalized medicine becomes more integrated into healthcare, the rate at which human genomes are being sequenced is rising quickly together with a concomitant acceleration in compute and storage requirements. To achieve the most effective solution for genomic workloads without re-architecting the industry-standard software, we performed a rigorous analysis of usage statistics, benchmarks and available technologies to design a system for maximum throughput. We share our experiences designing a system optimized for the "Genome Analysis ToolKit (GATK) Best Practices" whole genome DNA and RNA pipeline based on an evaluation of compute, workload and I/O characteristics. The characteristics of genomic-based workloads are vastly different from those of traditional HPC workloads, requiring different configurations of the scheduler and the I/O subsystem to achieve reliability, performance and scalability. By understanding how our researchers and clinicians work, we were able to employ techniques not only to speed up their workflow yielding improved and repeatable performance, but also to make more efficient use of storage and compute resources.

Dissecting force interactions in cellulose deconstruction reveals the required solvent versatility for overcoming biomass recalcitrance.

Pretreatment for deconstructing the multifaceted interaction network in crystalline cellulose is a limiting step in making fuels from lignocellulosic biomass. Not soluble in water and most organic solvents, cellulose was found to dissolve in certain classes of ionic liquids (ILs). To elucidate the underlying mechanisms, we simulated cellulose deconstruction by peeling off an 11-residue glucan chain from a cellulose microfibril and computed the free-energy profile in water and in 1-butyl-3-methylimidazolium chloride (BmimCl) IL. For this deconstruction process, the calculated free-energy cost/reduction in water/BmimCl is ∼2 kcal/mol per glucose residue, respectively. To unravel the molecular origin of solvent-induced differences, we devised a coarse graining scheme to dissect force interactions in simulation models by a force-matching method. The results establish that solvent-glucan interactions are dependent on the deconstruction state of cellulose. Water couples to the hydroxyl and side-chain groups of glucose residues more strongly in the peeled-off state but lacks driving forces to interact with sugar rings and linker oxygens. Conversely, BmimCl demonstrates versatility in targeting glucose residues in cellulose. Anions strongly interact with hydroxyl groups, and the coupling of cations to side chains and linker oxygens is stronger in the peeled-off state. Other than enhancing anion-hydroxyl group coupling, coarse-grain analysis of force interactions identifies configuring cations to target side chains and linker oxygens as a useful design strategy for pretreatment ILs. Furthermore, the state dependence of solvent-glucan interactions highlights specific stabilization and/or frustration of the different structure states of cellulose as important design parameters for pretreatment solvents.

Fabrication of protein-anchoring surface by modification of SiO2 with liposomal bilayer.

SiO(2) surface was successfully modified with phospholipid bilayer for biocompatibility by covering the planar surface with vesicular liposomes. By applying heat to rupture the vesicle, they were converted into a planar form. To effectively decorate the bilayer with biological molecules such as a protein, BAM (biological anchor for membrane) was used as a linker. It is a linear assembly consisting of oleyl chain, polyethylene glycol, and NHS (N-hydroxysuccinimide). After a target protein (BSA) was conjugated with BAM by NHS replacement, the conjugate was effectively inserted to the phospholipid bilayer through the lipophilic interaction between the oleyl chain and the lipid bilayer. The entire process was monitored and quantitatively analyzed by QCM (quartz crystal microbalance). BSA-BAM conjugate showed approximately 12-fold higher binding efficiency to the lipid bilayer than BSA alone. From this result, we conclude that SiO(2) surface could be modified to a lipid bilayer surface so as to anchor a protein by the action of BAM.

Inversion of radial distribution functions to pair forces by solving the Yvon-Born-Green equation iteratively.

We develop a new method to invert the target profiles of radial distribution functions (RDFs) to the pair forces between particles. The target profiles of RDFs can be obtained from all-atom molecular dynamics (MD) simulations or experiments and the inverted pair forces can be used in molecular simulations at a coarse-grained (CG) scale. Our method is based on a variational principle that determines the mean forces between CG sites after integrating out the unwanted degrees of freedom. The solution of this variational principle has been shown to correspond to the Yvon-Born-Green (YBG) equation [Noid et al., J. Phys. Chem. B 111, 4116 (2007)]. To invert RDFs, we solve the YBG equation iteratively by running a CG MD simulation at each step of iteration. A novelty of the iterative-YBG method is that during iteration, CG forces are updated according to the YBG equation without imposing any approximation as is required by other methods. As a result, only three to ten iterations are required to achieve convergence for all cases tested in this work. Furthermore, we show that not only are the target RDFs reproduced by the iterative solution; the profiles of the three-body correlation function in the YBG equation computed from all-atom and CG simulations also have a better agreement. The iterative-YBG method is applied to compute the CG forces of four molecular liquids to illustrate its efficiency and robustness: water, ethane, ethanol, and a water/methanol mixture. Using the resulting CG forces, all of the target RDFs observed in all-atom MD simulations are reproduced. We also show that the iterative-YBG method can be applied with a virial constraint to expand the representability of a CG force field. The iterative-YBG method thus provides a general and robust framework for computing CG forces from RDFs and could be systematically generalized to go beyond pairwise forces and to include higher-body interactions in a CG force field by applying the aforementioned variational principle to derive the corresponding YBG equation for iterative solution.

Intracellular amyloid beta interacts with SOD1 and impairs the enzymatic activity of SOD1: implications for the pathogenesis of amyotrophic lateral sclerosis.

Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease caused by the degeneration of motor neurons. Mutations in Cu/Zn superoxide dismutase (SOD1), including G93A, were reportedly linked to familial ALS. SOD1 is a key antioxidant enzyme, and is also one of the major targets for oxidative damage in the brains of patients suffering from Alzheimers disease (AD). Several lines of evidence suggest that intracellular amyloid beta (Abeta) is associated with the pathogenesis of AD. In this report we demonstrate that intracellular Abeta directly interacts with SOD1, and that this interaction decreases the enzymatic activity of the enzyme. We observed Abeta-SOD1 aggregates in the perinuclear region of H4 cells, and mapped the SOD1 binding region to Abeta amino acids 26-42. Interestingly, intracellular Ab binds to the SOD1 G93A mutant with greater affinity than to wild-type SOD1. This resulted in considerably less mutant enzymatic activity. Our study implicates a potential role for Abeta in the development of ALS by interacting with the SOD1 G93A mutant.

Fluorescence-based peptide screening using ligand peptides directly conjugated to a thiolated glass surface.

Functional peptides from peptide libraries are frequently screened using an array format. We report here results of a feasibility study of fluorescence-based peptide screening using an array format on surface-modified glass. The surface of an amine-coated glass slide was modified to contain thiol groups by iminothiolane treatment. The epsilon-amine of the C-terminal lysine from a ligand peptide was iodinated and then spotted onto the thiolated glass surface to covalently conjugate the ligand peptide to the surface via a thioether bond. This covalent immobilization allowed the ligand peptides to withstand washing steps by tightly adhering to the glass surface and confining their subsequent binding reactions within a spotted area. Two representative peptides were used as the ligand peptides; a 'target' (positive) heptapeptide that could specifically bind to trypsin, and a 'control' (negative) hexapeptide that had no binding affinity with trypsin. When fluorescein isothiocyanate-labeled trypsin was reacted with the ligand peptides, the target peptide demonstrated distinctively higher (ca. 8.7-fold) fluorescence intensity that was easily differentiated from the control peptide by a fluorescence scanner. A separate experiment using a quartz crystal microbalance confirmed that the difference in binding mass (ca. 9.1-fold) was very close to that seen in fluorescence intensity. These results suggested a quantitative, 1:1 correlation between mass and fluorescence signals. Furthermore, a smaller spot volume and a higher ligand peptide concentration resulted in higher fluorescence signal intensity. This study provides information on the potential for using fluorescence-based screening of functional peptides on a glass array format.